Re-use of this resource is governed by a Creative Commons
Attribution-
NonCommercial-ShareAlike 4.0 International
https://creativecommons.org/licenses/by-nc-sa/4.0/
The Blech
experiment
The Blech experiment illustrates the relationship between interconnect
length and the rate of electromigration-induced drift.
The test structure consists of aluminium (Al) segments with varying length
on the same underlying titanuim nitride (TiN) layer. These test structures
allow for a direct measurement of the electromigration displacement with
time.
When passing an electrical current through such a structure, the low resistivity
Al will drift on top of the TiN in the direction of electron flow - material
accumulation at the anode and material depletion at the cathode end.
Test structure
The TiN layer is stacked onto a SiO2 substrate. It is able
to carry an electrical current as it is a conductor. TiN has a low electrical
conductivity compared to Al.
Apply potential difference
An Al segment is then stacked above the TiN layer. As Al
is a better conductor than TiN, it is more favourable for the electric
current to flow through the Al segment.
Apply potential difference
The Blech
length
During electromigration, a stress gradient develops within the Al segment.
This is a result of mass atomic motion, resulting in regions of hydrostatic
tension and compression.
The critical stress gradient, which can be withstood before
failure, differs for Al segments of different lengths, following the relationship: (jl)c = ΔσΩ/Z*eρ
The stress gradient increases until it reaches a critical
limit. If the stress gradient does not exceed this critical value, no
damage will be observed.
Therefore, the shorter line will have infinite life, as
its critical stress gradient can never be achieved for the given current
density.
Electromigration
Damage
If the Al segment exceeds the critical length (Blech length), electromigration-induced
damage will be observed.
The damage is observed as a formation of voids and hillocks.